Major depressive disorder (MDD) is a common psychiatric illness with high cost to society and individual patients. One reason for the high cost is that most patients endure lengthy and ultimately unsuccessful empiric antidepressant trials before a beneficial medication is identified by trial-and-error. Care would be improved if a biomarker could determine, early in the course of treatment, whether a particular antidepressant would likely lead to response, remission, or treatment failure. Physicians could rapidly change treatments to an antidepressant which the biomarker indicated would be likely to help the patient. We propose a blinded and controlled feasibility study to evaluate a practical biomarker for predicting outcome based on data from the first week of antidepressant treatment. We have identified quantitative electroencephalographic (QEEG) changes that emerge early in the course of treatment with selective serotonin reuptake inhibitors (SSRIs) that appear to predict later response and remission. In this project, entitled PRISE-MD ("Personalized Response Indicators of SSRI Effectiveness in Major Depression"), we propose to move from biomarker discovery to the development phase with a prospective, controlled trial using treatments assigned based on biomarker status. Our specific aims are (1) to examine the utility of prospectively assigning antidepressant treatment using a QEEG-based biomarker guided treatment model, and (2) to evaluate the enhancement to the predictive accuracy of the model by including clinical, socio-demographic, or genetic factors. We will test specific hypotheses: (1) QEEG- biomarker-guided SSRI treatment will yield better rates of response than treatment irrespective of biomarker status;and (2) the predictive accuracy of the model will be enhanced by including clinical, socio-demographic, and genetic predictors. A total of 172 adults with MDD will receive a one-week pharmacologic challenge with the SSRI escitalopram (ESC), and then be assigned to double-blind treatment via stratified randomization: half of subjects with a positive biomarker (i.e., predicted to remit with ESC) will continue on ESC, and half will receive bupropion XL (BUP), a representative non-SSRI antidepressant which is a clinically-reasonable alternative. Similarly, half of subjects with a negative biomarker (i.e., predicted non-response with ESC) will continue with ESC and half will receive BUP. The co-primary outcomes will be 7-week response (50% improvement) on the 17-item Hamilton Depression Rating Scale and the 30-item Inventory of Depressive Symptomatology in subjects receiving the SSRI used for the biomarker (ESC). We will test our hypotheses (1) by examining outcomes from biomarker-guided treatment, and (2) by examining improvement in the prediction model from incorporating non-physiologic factors. A Data Safety Monitoring Board will monitor the utility of the biomarker in predicting treatment response and will have authority to halt the trial early if the biomarker proves to be a highly accurate predictor of response or non-response to ESC. This preliminary evaluation of applying our model to treatment selection will support further developments in the personalization of care for MDD. Public Health Relevance: Major depressive disorder (MDD) is a common psychiatric illness with high cost to society and individual patients worldwide, yet treatments chosen under current best practices frequently do not lead to recovery with the initial medication tried;this yields prolonged symptomatic suffering, functional disability, increased risk of relapse, and risk that individuals will abandon treatment efforts altogether. Better outcomes might be possible if a biomarker could guide clinicians in selecting among treatments. This project will examine biomarker predictions of outcome based on quantitative electroencephalographic (QEEG) features that change during a week of exposure to an SSRI antidepressant medication;by allowing clinicians to use treatments more effectively, the use of physiologic biomarker information for guidance could have a significant impact on the management of MDD.